摘要
为提高滚动轴承故障诊断识别准确率和计算效率,提出了一种基于改进SSA-VMD和高效样本熵的滚动轴承故障特征提取方法,并结合SVM进行诊断识别。首先,基于VMD算法中的模态数K和惩罚因子α是影响分解效果的两个关键因素,采用改进的SSA-VMD算法,以包络熵与峭度值比值的极小值为适应度值对VMD参数进行优化,并利用优化后的VMD进行信号分解及信号重构。其次,为进一步缩短特征提取的计算时间,引进结合率的思想,对统计不同维度下向量间距离小于阀值的向量个数时存在的重复计算步骤进行合并简化,得到一种高效的样本熵算法,利用高效样本熵对重构的信号构建特征向量。实验验证表明,上述方法不但提高了滚动轴承的故障诊断率,并且提高了故障诊断过程的计算效率。
In order to improve the accuracy and calculation efficiency of rolling bearing fault diagnosis and identi-fication,a rolling bearing fault feature extraction method based on improved SSA-VMD and high-efficiency sample entropy is proposed,and combined with SVM for diagnosis and identification.First,based on the number of modes K and penalty factorαin the VMD algorithm which are two key factors that affect the decomposition effect,the improved SSA-VMD algorithm is adopted,and the minimum value of the ratio of envelope entropy to kurtosis is the fitness val-ue.Optimize the VMD parameters,and use the optimized VMD for signal decomposition and signal reconstruction.Secondly,in order to further shorten the calculation time of feature extraction,the idea of combining rate is introduced,and the repeated calculation steps that exist when the distance between vectors in different dimensions are counted are less than the threshold value are combined and simplified,and an efficient sample entropy algorithm is obtained.,Use high-efficiency sample entropy to construct a feature vector for the reconstructed signal.Experimental verification shows that the method in this paper not only improves the fault diagnosis rate of rolling bearings but also improves the calculation efficiency of the fault diagnosis process.
作者
李彦阳
罗伟
LI Yan-yang;LUO Wei(College of Civil Engineering and Water Conservancy Institute,Heilongjiang Bayi Agricultural University,Daqing Heilongjiang 163319,China;College of Mechanical Science and Engineering,Northeast Petroleum University,Daqing Heilongjiang 163318,China;College of Intelligent Manufacturing,Hunan Railway Professional Technology College,Zhuzhou Hunan 412001,China)
出处
《计算机仿真》
2025年第9期269-273,573,共6页
Computer Simulation
基金
湖南铁道职业技术学院科研创新团队建设资助(KYTD202103)。
关键词
滚动轴承
高效样本熵
信号重构
故障诊断
Rolling bearing
Efficient sample entropy
Signal reconstruction
Fault diagnosis
